Most companies think they are “using AI” because a few employees have ChatGPT open in a browser tab. They ask it to draft emails, summarise documents, or brainstorm ideas. That is useful. But it is not what the next wave of AI is about.
Agentic AI is a fundamentally different category. Instead of answering questions when prompted, agentic systems perceive context, make decisions, and take actions autonomously. They process incoming data, update records, trigger workflows, and loop back to evaluate their own results. All without someone typing a prompt for every step.
The difference matters because the business impact is not incremental. It is structural.
ChatGPT is a conversation. An agent is a worker.
Think of ChatGPT as a very smart colleague you can ask questions. You type, it responds. It does not remember what happened yesterday unless you remind it. It cannot check your CRM, update a spreadsheet, or send a follow-up email on its own. Every interaction starts and ends with you.
An AI agent is different. You define a goal: “qualify every inbound lead within five minutes and route hot leads to sales.” The agent takes it from there. It reads the incoming message, checks the company against your ICP criteria, looks up firmographic data, scores the lead, drafts a reply, and either sends it or queues it for human review. Then it moves on to the next one.
The shift is from a tool you use to a system that works. IBM described it well: agents do not just follow instructions. They interpret goals and devise strategies.
Where the market stands today.
Agentic AI is not science fiction. 79% of organisations say they have adopted AI agents to some extent, according to PwC. McKinsey reports that 23% are scaling agentic AI systems, with another 39% experimenting.
But the numbers deserve closer reading:
- In any given business function, no more than 10% of respondents say their organisation is scaling AI agents
- Only about 6% of companies qualify as AI high performers (achieving more than 5% EBIT impact from AI)
- Those high performers are at least three times more likely than peers to be scaling agentic AI
- Only one in five companies has a mature governance model for autonomous AI agents
Translation: most companies are dabbling. A small group is pulling ahead. The gap between the two groups is widening.
Where agentic AI actually works today.
Not every process benefits from an autonomous agent. The use cases with the clearest ROI right now share common characteristics: they are repetitive, data-rich, time-sensitive, and involve clear decision rules.
Customer operations. One Harvard Business Review case study documented a company scaling AI-handled customer interactions from under 3% to nearly 60% within six months. The agent handles initial triage, resolves common requests, and escalates complex cases to humans with full context attached.
Sales qualification. AI sales agents continuously analyse customer data, past interactions, and outcomes to qualify leads, book meetings, and follow up automatically. They do not replace sales teams. They remove the manual work that keeps salespeople away from actual selling.
Process automation. 71% of organisations deploying AI agents do so specifically for process automation: invoice processing, data entry, compliance checks, report generation. These are high-volume, rule-based tasks where an agent can work 24 hours a day without fatigue or error drift.
Back-office operations. Research shows the biggest ROI from AI comes in back-office automation, not customer-facing tools. Yet over half of generative AI spending still targets sales and marketing. The opportunity in operations, finance, and procurement remains largely untapped.
The Dutch context.
The Netherlands is well positioned for agentic AI adoption. 81% of Dutch SMEs already operate in cloud environments. Digital infrastructure is strong. English proficiency means access to the latest tools and documentation is not a barrier.
But the adoption-value gap applies here too. Companies are buying AI tools without the organisational readiness to deploy them as agents. Data is siloed. Processes are not documented well enough for an agent to follow. Governance frameworks do not account for autonomous decision-making.
The EU AI Act adds another dimension. As of February 2025, AI literacy obligations are already in force. Full compliance is required by August 2026. Organisations deploying autonomous agents will need clear documentation of what those agents do, what data they access, and how humans maintain oversight. Starting that work now is not premature. It is necessary.
When to invest, and when to wait.
Agentic AI is not the right move for every company or every process. Here is a practical framework for deciding:
Invest now if:
- You have a clearly defined, repetitive process that consumes significant employee time
- Your data for that process is clean, structured, and accessible via APIs
- The decision rules are well understood (even if complex)
- You can define what “good output” looks like and measure it
- You have someone who can own the system once it is running
Wait if:
- Your underlying processes are not documented or standardised
- Data lives in disconnected systems with no integration layer
- The task requires nuanced judgment that changes frequently
- You do not have clear metrics for what success looks like
Waiting is not the same as doing nothing. If you are not ready for agentic AI, the preparation work (process documentation, data cleanup, integration) will pay off regardless. It makes every future technology decision easier.
The market is moving fast.
Gartner forecasts that 40% of enterprise applications will embed task-specific AI agents by 2026, up from less than 5% in 2025. By 2028, at least 15% of day-to-day work decisions will be made autonomously through agentic AI. The AI agent market itself is projected to grow from EUR 7.8 billion in 2025 to EUR 52.6 billion by 2030.
For mid-market companies, this means the window for early-mover advantage is now. The companies that figure out agentic AI in the next 12 to 18 months will have compounding advantages over those that wait for the technology to become “easier.” It will get easier. But it will also get more competitive.
Start with the right question.
Do not start with “should we build an AI agent?” Start with “which of our processes burns the most time, follows clear rules, and runs on accessible data?” If you find a strong candidate, the technology is ready. If you do not, the preparation work is still worth doing.
At Practical North, we help companies answer exactly that question. Our North Star Audit maps your workflows against current AI capabilities, identifies where the investment is justified, and defines a concrete next step. Three hours. No retainer.
The gap between “we use ChatGPT” and “we have AI agents running our operations” is where the real competitive advantage lives. The question is which side of that gap you want to be on.